Wei Q. Iterative Adaptive Dynamic Programming..Self-Learning Optim. Control 2026

Textbook in PDF format

This unique book introduces the iterative adaptive dynamic programming theory from the control systems perspectives, with the most recent results of iterative adaptive dynamic programming methods. Advanced theoretical analysis and some practical applications of iterative adaptive dynamic programming are provided. Furthermore, the practical applications in residential energy systems are also highlighted, showing the good performance of the iterative adaptive dynamic programming methods.
Adaptive programming (ADP) employs approximation structures, such as neural networks, to approximate the performance index function in dynamic programming, such that the performance index function satisfies the HJB equation. There are generally three modules in ADP, which are model, critic and action modules, respectively. Every module can be realized by neural networks.
Dynamic programming often suffers from the curse of dimensionality, which is primarily due to the backward-in-time mechanism. To avoid the difficulty, based on function approximators, such as neural networks, adaptive dynamic programming is an effective brain-like method to solve optimal control problems forward-in-time. Recently, the study on adaptive dynamic programming and related fields has gained much attention from various scholars.
The useful reference text benefits professionals, researchers, academics, graduate and undergraduates students in control engineering